Lectures on Algebraic Groups

Dipendra Prasad
Notes by Shripad M. Garge
1. Basic Ane Algebraic Geometry
We begin these lectures with a review of ane algebraic geometry.
Let k be an algebraically closed eld. An Ane algebraic variety over k is
a subset X A
n
:= k
n
of the form
X =
_
x A
n
: f
i
(x) = 0 for certain f
1
, . . . , f
r
k[x
1
, . . . , x
n
]
_
.
Thus, an ane algebraic variety is the set of common zeros of certain poly-
nomial equations. The coordinate ring of an ane variety X, denoted by k[X],
is the ring of polynomial functions on X; it is given by
k[X] :=
k[x
1
, . . . , x
n
]
_
(f
1
, . . . , f
r
)
,
where for an ideal I, we dene

I =
_
g k[x
1
, . . . , x
n
] : g
m
I for some m 0
_
.
The notions of subvariety and a nite product of varieties make sense and
are dened in the most natural way.
If X A
n
, and Y A
m
are ane algebraic varieties, then by a polynomial
map from X to Y , we mean a mapping from X to Y which is the restriction to
X of a mapping from A
n
to A
m
given by
(x
1
, . . . , x
n
)
_

1
(x
1
, . . . , x
n
), . . . ,
m
(x
1
, . . . , x
n
)
_
1
with
i
k[x
1
, . . . , x
n
].
The map X k[X] denes a natural contravariant transformation. Clearly
a polynomial map f : X Y gives rise to a map from f

: k[Y ] k[X]
dened by, f

() = f.
X
f
//
f

()

@
@
@
@
@
@
@
@
Y

k
It is obvious that such a function on X is given by a polynomial function.
Theorem. There exists a functorial correspondence between ane algebraic
varieties and nitely generated k-algebras without nilpotents. This correspon-
dence is contravariant.
One denes a topology on an ane algebraic variety X by taking closed sets
of X to be zeros of polynomials. It is called as the Zariski topology on X.
If X
1
X and X
2
X are closed subsets then X
1
X
2
and X
1
X
2
are
closed. Indeed, if X
1
, X
2
are given by {f

}, {g

} respectively, then X
1
X
2
and
X
1
X
2
are given by {f

, g

} and {f

} respectively.
One important dierence between the Zariski topology and the usual metric
topology is that the Zariski topology is not Hausdor. This can be seen already
for X = A
1
. Here the closed sets are precisely all subsets of nite cardinality
besides the whole set X. Thus non-empty open sets are complements of nite
sets. Therefore any two non-empty open sets intersect, and hence the space A
1
is not Hausdor with the Zariski topology.
Proposition. Let X be an algebraic variety.
(i) Any family of closed subsets of X contains a minimal one.
(ii) If X
1
X
2
. . . is a descending sequence of closed subsets of X, then
there exists k such that X
i
= X
k
for all i k.
Proof. The proof follows as the polynomial algebra k[x
1
, . . . , x
n
] is Noetherian.
.
2
The property (i) in the above proposition states that an algebraic variety is
Noetherian with the Zariski topology. Note that the two properties in the above
proposition are equivalent!
An irreducible algebraic variety is the one in which any two non-empty open
sets intersect. Example: A
1
is irreducible.
An algebraic variety is irreducible if and only if any non-empty open set is
dense in it. A subvariety is irreducible if it is irreducible in the induced topology.
Following statements are easy to prove.
Lemma.
(i) A subvariety Y of an algebraic variety X is irreducible if and only if Y
is irreducible.
(ii) If : X Z is a morphism and X is irreducible, then so is (X).
(iii) If X
1
, X
2
are irreducible varieties, then so is X
1
X
2
.
Note that an irreducible algebraic variety is always connected but the converse
is not true. Example: Take X = {(x, y) A
2
: xy = 0}. Here the open sets,
viz., Y
1
= {(x, 0) X : x = 0} and Y
2
= {(0, y) X : y = 0} do not
intersect each other. And X = {(x, 0)} {(0, y)}, where both the sets {(x, 0)}
and {(0, y)} are connected (being homeomorphic to A
1
) and they intersect each
other, hence the union is connected.
The geometric intuition behind the irreducible variety is that a general va-
riety is a nite union of components where a component means a maximal
irreducible subset.
Irreducible varieties correspond to minimal elements in the primary decom-
position of the ideal (0) in the coordinate ring k[X]. A variety X is irreducible
if and only if the corresponding coordinate ring k[X] is an integral domain.
Let X be an irreducible variety. Let k(X) denote the eld of fractions of the
integral domain k[X]. This is the eld of meromorphic functions on the variety
X. We dene dimension of the irreducible variety X to be the transcendence
degree of k(X) over k. Dimension also equals the length of maximal chain of
irreducible subvarieties.
Examples:
3
(i) A
1
: {0} A
1
, dim(A
1
) = 1;
(i) A
2
: {0} A
1
A
2
, dim(A
2
) = 2.
For a general variety X, the dimension is maximal of dimensions of its irre-
ducible components.
Lemma. Let X, Y be irreducible varieties of dimensions m, n respectively,
then dimX Y = m+n.
Proof. This is clear since k[X Y ] = k[X] k[Y ]. .
Lemma. Let X be an irreducible variety and let Y be a proper subvariety of
X. Then dimY < dimX.
Proof. Let k[X] = k[x
1
, . . . , x
r
] and k[Y ] = k[X]/P, where P is a non-
zero prime ideal. Let y
i
be the image of x
i
in k[Y ]. Let dimX = m and
dimY = n. We can assume that y
1
, . . . , y
n
are algebraically independent. Then
clearly x
1
, . . . , x
n
are algebraically independent, hence m n. Assume that
m = n. Let f be a nonzero element of P. Then there is a nontrivial relation
g(f, x
1
, . . . , x
n
) where g(t
0
, . . . , t
n
) k[t
0
, . . . , t
n
].Since f = 0, we can assume
that t
0
does not divide all monomials of g, hence h(t
1
, . . . , t
n
) := g(0, t
1
, . . . , t
n
)
is nonzero, but then h(y
1
, . . . , y
n
) = 0 contradicting the algebraic independence
of y
i
. This completes the proof. .
A locally closed set is an open subset of a closed set. Example: A

A
2
as the subset {(x, 0) : x A

}, is locally closed.
A constructible set is a nite union of locally closed sets. Example: A
2

A
2
is constructible but not locally closed.
Theorem. (Chevalley). If f : X Y is a morphism of algebraic
varieties, then f(X) is constructible.
Example: Let f : A
2
A
2
be the morphism given by f(x, y) = (x, xy). Then
f(A
2
) = A
2
A

.
We shall also need the notion of projective varieties. These are the varieties
that are dened by homogeneous polynomials. These varieties can be seen as
closed subsets of P
n
for some n, hence the name projective.
2. Ane Algebraic Groups
An ane algebraic group G is an ane algebraic variety as well as a group
4
such that the maps m : G G G and i : G G, given by m(x, y) =
xy, i(x) = x
1
, are morphisms of algebraic varieties.
Examples of ane algebraic groups:
(i) Any nite group can be made into an algebraic group. To make G into an
algebraic group, we have to give a nitely generated k-algebra k[G]. We
take k[G] as the algebra of functions on G.
(ii) The groups GL
n
, SL
n
, Sp
2n
, SO
n
, O
n
, U
n
, etc. are some of the standard
examples of ane algebraic groups.
For the group GL
n
, the structure of an ane algebraic group is given by,
GL
n
=
__
X 0
0 x
n+1
_
: det X x
n+1
= 1
_
.
The multiplication m : GL
n
GL
n
GL
n
is a polynomial map, (X)
ij
(Y )
ij

(Z)
ij
. To say that the map m is polynomial is equivalent to say that coordinates
of Z depend on those of X and Y polynomially.
As a special case, for n = 1, GL
1
= G
m
= k

and the coordinate ring is

k[GL
1
] = k[x][x
1
].
Lemma. A closed subgroup of an algebraic group is an algebraic group.
Proof. Clear from the denitions.
Remark. A locally closed subgroup is closed. In fact every open subgroup is
closed. (Hint: Coset decomposition)
This remark is not true in the case of topological groups. Example: A line
with irrational slope in R
2
, gives an embedding of R into R
2
/Z
2
as an everywhere
dense subgroup of the torus R
2
/Z
2
.
Lemma. Let : G
1
G
2
be a homomorphism of algebraic groups then
(G
1
) is a closed subgroup of G
2
.
Proof. From Chevalleys theorem, (G
1
) is constructible. We can assume
without loss of generality, that (G
1
) is dense in G
2
.
Since (G
1
) is constructible, it contains an open subset of G
2
, hence (G
1
)
is open in G
2
. And then by previous remark (G
1
) = G
2
. This completes the
proof.
Lemma. A connected algebraic group is irreducible.
5
Proof. One needs to prove that there is a unique irreducible component of
G passing through {e}, the identity of the group G. Let X
1
, . . . , X
m
be all
irreducible components of G passing through {e}.
Look at the mapping : X
1
X
m
G, given by multiplication.
Since X
i
are irreducible, so is their product and also the image of the product in
G under the map . Clearly the image contains identity and therefore the image
of the map is contained in an irreducible component of G, say X
1
. Since all
X
i
contain identity, this implies that all X
i
are contained in a xed X
1
.
Lemma. The irreducible component of G passing through {e} is a closed
normal subgroup of G of nite index.
Proof. We denote the irreducible component of G passing through {e} by G

.
Being closed is a general property of irreducible algebraic varieties, hence G

is
closed in G. To prove that G

is a subgroup of G, we must show that whenever

x, y G

, xy
1
G

. Clearly x
1
G

is also irreducible and e x

1
G

,
thus x
1
G

= G

, i.e., x
1
y G

x, y G

. Similarly one can prove that

G

is normal.
An algebraic variety has nitely many irreducible components, hence G

is of
nite index in G. This completes the proof.
Lemma. For an algebraic group G, any closed subgroup of nite index con-
tains G

.
Proof. Let H be a subgroup of G of nite index. Then H

is closed and hence

open in G

(being a subgroup of nite index), hence H

= G

.
Remark. If n is odd, then there is no real dierence between SO(n) and O(n)
as O(n) = {1} SO(n).
0 SO(n) O(n) Z/2 1
0 (Z/2)
n1
W(SO(2n)) S
n
1
0 (Z/2)
n
W(SO(2n + 1)) S
n
1
3. Group Actions on Algebraic Varieties
Let G be an algebraic group, and X an algebraic variety. A group action
of G on X is a morphism of algebraic varieties : G X X such that
(g
1
, (g
2
, x)) = (g
1
g
2
, x) and (e, x) = x, g
1
, g
2
G, x X.
6
An algebraic variety X is called homogeneous if G operates transitively on
X. Isotropy subgroups, orbits are dened in the natural way.
Lemma. Isotropy subgroups are closed. Orbits are open in their closure.
Proof. The map : G X X gives rise to a map
1
: G X X
dened by, g

1
(x
0
, gx
0
) for a xed x
0
X. Then the isotropy subgroup
of x
0
, viz. G
x
0
is the inverse image of (x
0
, x
0
) under the map
1
and hence is
closed.
G operates on X, Gx
0
X. In fact, Gx
0
Gx
0
X. The set Gx
0
is
constructible, i.e., contains an open subset U of Gx
0
, and Gx
0
is union of the
cosets gU, g Gx
0
, therefore Gx
0
is open in Gx
0
.
Lemma. For any group action : G X X, there is always a closed
orbit in X.
Proof. It suces to prove the lemma under the condition that G is irreducible
because if Z is a closed orbit under the action of G

, then G Z is a nite union

of closed sets, and hence is closed, and Z being an orbit of G

, G Z is an orbit
of G. So, we now assume that G is irreducible.
Choose an orbit whose closure has smallest dimension, say Gx
0
. Write Gx
0
=
Gx
0
Y . As Gx
0
is G invariant, so is Gx
0
and hence Y is also G invariant,
then dimY < dimGx
0
and Y is closed. Contradiction!
Some standard contexts for group action:
(i) G acts on itself via left and right translations, and this gives rise to an
action of GG on G, given by, (g
1
, g
2
) g = g
1
gg
1
2
. This is a transitive
action, the isotropy subgroup of identity being G G.
(ii) If V is a representation of G, then this gives an action of G on V , and one
may classify orbits on V under this action.
Exercise 1. The group G = GL
n
operates on V = C
n
in the natural way.
Classify orbits of G action on Sym
2
(V ) and
2
(V ) induced by this natural action.
Answer. Action of GL
n
on Sym
2
(V ) is essentially X gX
t
g. Number of
orbits in V is 2, in Sym
2
(V ) it is n + 1 and in
2
(V ) it is
_
n+1
2

.
Our main aim in this section is to prove that any ane algebraic group is
linear, i.e., it is a closed subgroup of GL
n
for some n. The proof depends on
7
group actions on algebraic varieties. We will be dealing exclusively with ane
algebraic varieties. The coordinate ring of an ane algebraic group G is denoted
by k[G], which is same as the space of regular functions on G.
Note that k[GX]

= k[G] k[X], i.e., polynomial functions on the product

space G X are a nite linear combination of functions of the form p(g)q(x)
where p is a polynomial function on G and q is a polynomial function on X.
This crucial property about polynomials goes wrong for almost any other kind of
functions.
Exercise 2. Show that cos(xy) cannot be written as a nite linear combination
of functions in x and y alone.
Answer. If cos(xy) were a nite sum:
cos(xy) =

f
i
(x)g
i
(y),
it follows by specialising the y value, that cos(ax) is a linear combination of
nitely many functions f
i
(x) for all a, i.e., the space of functions cos(ax) gen-
erates a nite dimensional vector space. But this is false, as for any n, given
distinct a
1
, . . . , a
n
0, cos(a
i
x) are linearly independent (Hint: Van der monde
determinant).
Proposition. If X is an ane algebraic variety on which G operates, then
any nite dimensional space of functions on X is contained in a nite di-
mensional space of functions which is invariant under G.
Proof. It suces to prove that the translates of a single function is nite
dimensional.
Consider : G X X. This map gives rise to a ring homomorphism

: k[X] k[GX] = k[G] k[X], given by, f

i
f
i
,
i.e., f(gx) =

i
(g)f
i
(x),
i.e., f
g
=

i
(g)f
i
and hence f
g
span of f
i
g G.
Proposition. If X is ane algebraic variety on which G acts, then a sub-
space F of k[X] is G-invariant if and only if

(F) k[G] F, where

is
as in previous proposition.
Proof. If

(F) k[G] F, then there are functions

i
on k[G] such that
f
g
(x) =

i
(g)f
i
(x) for f
i
F. So, f
g
F, i.e., F is G-invariant.
8
To prove the converse, suppose F is G-invariant and let {f
r
} be a basis
of F. We extend this basis to a basis of k[X] by adjoining say {g
s
}. Let
f F. Then,

(f) k[G] k[X]. This implies that

(f) =

r
a
r

f
r
+

s
b
s
g
s
, for certain polynomial functions {a
r
}, {b
s
} on G. Therefore
f(gx) =

r
a
r
(g)f
r
(x) +

s
b
s
(g)g
s
(x). But since F is G-invariant, all b
s
are
identically zero and hence

(F) k[G] F.
Theorem. Any ane algebraic group is linear, i.e., it is isomorphic to a
closed subgroup of GL
n
for some n.
Proof. The coordinate ring of G, k[G] is a nitely generated k-algebra, i.e.,
there exist functions f
1
, . . . , f
r
which generate k[G] as an algebra over k. By
proposition, one can assume that the subspace generated by f
i
s is G-invariant
(G acts by left translation).
Thus, f
i
(gx) =

m
ij
(g)f
j
(x) x G, i.e., f
g
i
=

m
ij
(g)f
j
As

(F) k[G] F, so m
ij
can be assumed to be algebraic functions on
G. This gives rise to a map : G GL
n
(k), given by g (m
ij
(g)), which
is a homomorphism of algebraic groups. We will show that this identies G as a
closed subgroup of GL
n
. Since the image of an arbitrary map of algebraic groups
is closed, the image of G, call it H, is a closed subgroup of GL
n
.
We will prove that the surjective mapping : G H is an isomorphism
of algebraic groups. For this, it suces to prove that the induced map on the
coordinate rings

: k[H] k[G] is a surjection. But this follows since f

i
are
in the image of k[H].
This completes the proof.
Remark. Note that a bijective map need not be an isomorphism of algebraic
varieties, as the example x x
p
on A
1
in characteristic p shows. As another
example, consider the map from A
1
to A
2
given by x (x
2
, x
3
). This gives a
set-theoretic isomorphism of A
1
into the subvariety {X
3
= Y
2
} A
2
, but it is
not an isomorphism of algebraic varieties. (Why?)
A morphism f : X Y is called dominant if f(X) is dense in Y ; Example:
the map f : A
2
A
2
dened by, (g
1
, g
2
) (g
1
, g
1
g
2
).
It can be seen that f is dominant if and only if f

: k[Y ] k[X] is injective.

It can be seen that f is an isomorphism if and only if f

is so.
Exercise 3. Classify continuous functions f : R C such that the span of
9
translates of f is nite dimensional.
Answer. It is easy to see that the functions f on R such that the span of
translates of f is nite dimensional is an algebra, i.e., the set of such functions
is closed under addition and multiplication. Any polynomial has this property
and so do the functions e
x
. We shall prove that polynomials and exponential
functions generate the space of such functions as an algebra.
Let f be a continuous complex valued function on R, and let V be the nite
dimensional space spanned by the translates of f. If {f
1
, . . . , f
n
} is a basis of
V , then
(t)(f
i
)(x) = f
i
(t +x) =

g
ij
(t)f
j
(x) t, x.
Then t g
ij
(t) is a matrix coecient of the nite dimensional representation
V of R.
Putting x = 0, we get
f
i
(t) =

g
ij
(t)f
j
(0).
So f
i
are sum of matrix coecients. Thus, it suces to understand matrix
coecients of nite dimensional representations. So, we come upon a question,
that of classifying nite dimensional continuous representations of R.
Claim. Any nite dimensional representation of R is of the form t e
tA
for
some matrix A M
n
(C).
If the representation was analytic, then this follows from Lie algebra methods.
However any continuous representation is analytic. We also give another way to
characterise nite dimensional representations of R
We write g(t) = log(f(t)) in a neighbourhood of 0. Then, g is a map from R
to M
n
(C) such that t
1
+t
2
g(t
1
)+g(t
2
). As any continuous homomorphism
from R to itself is a scalar multiplication, g(t) = tA for some matrix A. Then
f(t) = exp(tA) in a neighbourhood of identity and a neighbourhood of identity
generates all of R, so the claim is proved.
Then by canonical form of A,
e
tA
=

e
t
i
exp(tN
i
)
and exp(tN
i
) is a polynomial in t, as N
i
are nilpotent. This answers the question.
4. Some Generalities about Closures in the Zariski Topology.
10
Given A X, where X is an algebraic variety, one can dene A to be the
smallest closed algebraic subvariety of X containing A, i.e.,
A =

{Y : A Y, Y is closed in X}.
In particular, if G an algebraic group and H is an abstract subgroup of G, one
can talk about H which is a closed subvariety of G.
Lemma. If H is an abstract subgroup of G, then H is a closed algebraic
subgroup of G.
Proof. We need to prove H H H and H
1
H.
Clearly, H h
1
H for any h H and h
1
H is also closed in G.
Hence, H h
1
H
h H H h H
H H H
H h H h H
H h H h H
H H H.
Similarly by noting that x x
1
is a homeomorphism of G, one can prove
that H is closed under inversion. Thus H is a closed subgroup of the algebraic
group G.
This group H is called the algebraic hull of H.
Proposition. If G GL
n
(C) is a subgroup such that for some e 1,
x
e
= 1 x G, then G is nite.
Proof. If G is not nite, look at G, an algebraic subgroup of GL
n
(C), for
which x
e
= 1 continues to hold good. There exists a subgroup of G of nite
index, G

, such that G

is connected. For a connected algebraic subgroup G

,
G

(C) is a Lie group of positive dimension, and then x

e
can not be identically
1. Contradiction!
Let H GL
n
(C). Thinking of GL
n
(C) as an algebraic group dened over
Q, one can talk about closure of H as a subgroup of GL
n
in the Zariski topology
over C or the one over Q. We have
H =

{Y : H Y, Y is closed}.
11
If we take only those Y , which are dened over Q, we get the closure of H over
Q, which we denote by H
Q
. Clearly H
Q
H.
Example: Since e is transcendental over Q, the only polynomial f Q[x] with
f(e) = 0 is the zero polynomial. Hence e
Q
= C.
Exercise 4. Let G be an algebraic group. Then < g > is Abelian. However, if
G = SL
n
(C), there are elements whose Zariski closure in Q-topology is SL
n
(C),
so non-Abelian.
Answer.
Mumford-Tate group.
For a smooth Abelian variety X, H
1
(X) with coecients in Q, admits a
decomposition over C, called as Hodge decomposition as,
H
1
(X)
Q
C = H
0,1
H
1,0
.
One denes a subgroup of GL(H
1
) which is Zariski closure in the Q-topology
of the subgroup
_

_
_

_
z
.
.
.
z
z
.
.
.
z
_

_
: z C

_
.
This subgroup is called as the Mumford-Tate group associated to X.
Examples:
(i) If X is a CM elliptic curve, then MT(X) = k

.
(ii) If X is a non-CM elliptic curve, then MT(X) = GL
2
.
(iii) If X is a Klein curve, then MT(X) = k

.
Tannakas Theorem. A closed subgroup of SO(n, R) (in the Euclidean
topology of R) is the set of real points of an algebraic group dened over R.
Proof. Let G be a closed subgroup of SO(n, R). Any polynomial over C can
be written as f = f
1
+if
2
where f
i
are polynomials dened over R, then
f(z) = 0 if and only if f
1
(z) = 0 and f
2
(z) = 0.
12
for z SO(n, R), therefore G
C
= G
R
.
We want to prove that G = G(R). G is a closed subgroup of G(R) which is
a closed subgroup of GL
n
(R).
Suppose G G(R). Let g belong to G(R) but not to G. Since g G(R),
every polynomial vanishing on G also vanishes on g. If we can show the existence
of a polynomial vanishing on G but not on g, we will be done.
Look at disjoint closed sets G and gG in G(R). By Urysohns lemma, there
exists a continuous function f on G(R) which is 0 on G and 1 on gG. Observe
that the coordinate functions {m
ij
(g)} are continuous functions on G, and hence
by Stone-Weierstrass theorem, the algebra generated by m
ij
is dense in C(G(R)).
Therefore, there exists a polynomial in m
ij
, say p(m
ij
), which approximates f
very closely on G(R).
But this polynomial need not be identically zero on G which is what we want
to achieve. To this end, we dene
F(g) =
_
G
p(m
ij
)(gh)dh.
Since F is a right G-invariant function, it is constant on G and on gG. Thus
F =
1
on G and F =
2
on gG, where we can assume that
1
is very close to 0,
and
2
is very close to 1. Then by suitable translation, F is 0 on G and nonzero
at g. So, the only remaining thing to prove is that F is a polynomial again in
m
ij
.
We note the property of matrix coecients that
m
ij
(gh) =

k
m
ik
(g) m
kj
(h).
Therefore for any polynomial p in the n
2
-variables m
ij
, we have
p(m
ij
)(gh) =

P,Q
P(m
ij
)(g)Q(m
ij
)(h)
where P and Q are certain polynomials in m
ij
.
This completes the proof.
Exercise 5. If V is a faithful representation of a compact group G, then any
irreducible representation W is contained in V
r
V
s
.
13
Answer. Assume the contrary, then
_
g
W
f
V
= 0
for all matrix coecients g
W
of W and f
V
of V
r
V
s
. The C-sum of matrix
coecients of V
r
V
s
forms a subalgebra of functions on G. So
_
g
W
f = 0
for all g
W
and all f in this subalgebra. But by Stone-Weierstrass theorem, this
subalgebra is dense on G. Contradiction!
Corollary.
(i) If V is any faithful representation, then the algebra generated by the
matrix coecients of V is independent of V .
(ii) (Coro. of Exer. 3 and Exer. 5) The space of functions f in C[G] which
span a nite dimensional space under left (right) translations is pre-
cisely the algebra of matrix coecients of nite dimensional represen-
tations.
For a compact group G SO(n, R), we discussed about algebraic closure
of G, viz. G
alg
in R-Zariski topology such that G
alg
(R) = G. G
alg
is called the
complexication of G, in the sense that Lie(G) C = Lie(G
alg
)(C).
The above statement need not be true for non-compact groups. Torus creates
some problems. The torus R

has few algebraic characters, but many

topological characters.
Peter-Weyl Theorem.
(i) Any compact Lie group has a faithful representation.
(ii) L
2
(G)

V
V V

as GG module, where V runs over all irreducible

representations of G.
5. Lie Algebras associated to Algebraic Groups.
Let k be a ring, A a commutative k-algebra and M an A-module. A deriva-
tion d of A with values in M is a map d : A M, such that
14
d(a +b) = d(a) +d(b) a, b A
d(ab) = ad(b) +bd(a) a, b A
d(a) = d(a) k, a A.
(Observe that d(r) = 0 r k.) The space of derivations of A with values
in an A-module M is naturally an A-module. If M = A, then a derivation of A
with values in M is called a derivation on A.
Examples:
(i) A = k[x] = M, then any derivative is of the form d(g) = f(x)
d
dx
(g),
for some f k[x]. Hence, the space of derivations is free of rank 1 over
A = k[x].
(ii) A = k[x
1
, . . . , x
n
] = M, the derivations

x
i
form a free basis of the space
of derivations over A.
Exercise 6. Justify above statement.
Answer. Let d be a derivation. Dene d(x
i
) = f
i
. Now, we have two deriva-
tions d
1
=

f
i

x
i
and d, but they are same on generators, viz., x
i
, so they are
same on the full algebra k[x
1
, . . . , x
n
].
Corollary. The space of derivations on A = k[x
1
, . . . , x
n
] is a free A-module
of rank equal to the dimension of A over k.
Let O
X,x
be the ring of germs of functions at a point x X, where X is
either a manifold or an algebraic variety over a eld k.
In algebraic geometry, O
X,x
is the space of rational functions
f
g
, where f
and g are polynomial functions dened in a neighbourhood of the point x and
g(x) = 0, whereas in topology O
X,x
is the space of C

functions dened in a
neighbourhood of x.
A = O
X,x
, M = k = O
X,x
/m
x
, where m
x
:= ker{f f(x)}. A derivation
in this case is called a tangent vector at x X.
Let T : O
X,x
k be a derivation, i.e., T has the property that T(fg) =
f(x)T(g) + g(x)T(f). Then T(1) = T(1) + T(1). Therefore T(1) = 0. Also
T : m
x
k factors through m
x
2
, so it induces a map : m
x
/m
x
2
k. We
will prove that the tangent vectors at x X can be canonically identied to
(m
x
/m
x
2
)

. The space of tangent vectors at x X is denoted by T

X,x
.
15
Note that m
x
/m
x
2
is a nite dimensional k-vector space (follows from Noetherian-
ness and Nakayamas lemma). We have the inequality
dim
_
m
x
/m
x
2
_
dimX x X.
When equality holds, we say x is a smooth point.
In characteristic 0, the set of singular points is a proper closed subvariety. In
any characteristic, the Jacobian criterion is satised.
Jacobian Criterion. If f(x
1
, . . . , x
n
) = 0 the is equation of a variety, then
x is smooth if and only if there exists i such that
f
x
i
= 0 at x.
A derivation d =

f
i

x
i
is visualised as associating vectors (f
1
(x), . . . , f
n
(x))
to any point x.
Lemma.
(i) T
X,x

= Hom
k
(m
x
/m
x
2
, k).
(ii) When k = R, for any point x on an n-dimensional R-manifold X,
m
x
/m
x
2
is an n-dimensional R-vector space.
Proof. (i) Observe that d(m
x
2
) = 0, as d(fg) = f(x)dg + g(x)df and
f, g m
x
therefore f(x) = g(x) = 0. So d factors through m
x
2
, hence
d Hom
k
(m
x
/m
x
2
, k).
Conversely, we need to prove that a k-linear map T : m
x
/m
x
2
k gives
rise to a derivation.
We dene d(f) = T(f f(x)). Clearly d is linear. Now, d(fg) = T(fg
fg(x)) and fdg +gdf = fT(g g(x)) +gT(f f(x)). Hence to prove d(fg) =
f(x)dg +g(x)df, we need to prove
T(fg f(x)g(x) f(x)g +f(x)g(x) g(x)f +f(x)g(x)) = 0.
This follows as
fg f(x)g g(x)f +f(x)g(x) =
_
f f(x)
__
g g(x)
_
m
x
2
.
For (ii), we have dim(m
x
/m
x
2
) n as

x
i
are linearly independent over R.
To prove other inequality we use following fact.
16
Any f C

(R
n
) can be written locally around the origin as
f = f(0) +

f
i
X
i
+

g
ij
X
i
X
j
where f
i
are constants and g
ij
C

(R
n
).
Thus T
X,x
, the space of derivations on O
X,x
, is an n-dimensional vector space,
called the tangent space to X at x. A typical element of T
X,x
looks like

f
i

x
i
,
where f
i
are some constants.
A map, x v
x
T
X,x
, varying smoothly, is called a vector eld, i.e., if we
write v
x
=

f
i
(x)

x
i
, then f
i
are smooth functions.
One can also dene smoothness of a vector eld V by saying that V (f) is a
smooth function for all f smooth.
Given vector elds V
1
and V
2
, one denes another vector eld [V
1
, V
2
] =
V
1
V
2
V
2
V
1
. More precisely
[V
1
, V
2
](f) = V
1,x
(V
2
f) V
2,x
(V
1
f).
This can be checked to be a vector eld, called the Lie bracket. This denes a
Lie algebra structure on the set of vector elds on a manifold.
Functorial Properties of tangent spaces. If f : X Y is a morphism,
then we have a map df : T
X,x
T
Y,f(x)
given by, df(v)() = v( f).
Caution: One cannot use df to push vector elds from X to Y as there might
be several points in X with the same image in Y .
If V
1
is a vector eld on X and V
2
on Y , then V
2
is f-related to V
1
if, for all
x X, (df)
x
: T
X,x
T
Y,f(x)
sends V
1,x
to V
2,f(x)
.
In particular, if G is a group which operates on a manifold X in a dierentiable
way, it makes sense to talk about vector elds on X, invariant under G.
Examples:
(i) On R, any vector eld d looks like f(x)
d
dx
. The group R acts on itself by
translations and only vector eld invariant under addition is
d
dx
for a
constant.
(ii) On R

, x
d
dx
is the only invariant vector eld upto scalar multiple.
17
(iii) Consider the group GL
n
(R) M
n
(R). A vector eld on GL
n
(R) looks
like

a
ij
(X)

X
ij
. Any left-invariant vector eld corresponds to a matrix
(b
ij
) M
n
(R), and the invariant vector eld is

i,j,k
b
ik
X
kj

X
ji
.
Exercise 7. Justify above sentence. (Hint: Change of variables)
Proposition. If 0 X k
n
, where X is dened by polynomial equations
f
i
(x
1
, . . . , x
n
) = 0, then the tangent space to X at 0 is dened by degree 1
part of f
i
.
Proof. As 0 X, f
i
(0, . . . , 0) = 0. We write f
i
(x
1
, . . . , x
n
) =

+ . . . .
We have already noted that T
X,x

= (m
x
/m
x
2
)

. We note that if X is a closed

subvariety of A
n
, then the ring of polynomial functions on X, i.e., k[X] is a
quotient of k[x
1
, . . . , x
n
], hence derivations of k[X] at any point of X can be
identied to a subspace of the space {

x
1
, . . . ,

x
n
}. Hence
m
0
m
2
0

=
< x
1
, . . . , x
n
>
I+ < x
1
, . . . , x
n
>
2
.
Examples: 1. The curve X
2
= Y
3
at (0, 0) has 2-dimensional tangent space
at (0, 0) as there is no linear term. Similarly the tangent space to the curve
X
2
= Y
2
is 2-dimensional.
2. X = Y
4
+Z
7
has 2-dimensional tangent space given by

Y
,

Z
.
One can calculate Lie algebra associated to algebraic groups by using above
proposition.
O(n) = {A GL
n
:
t
AA = I} M
n
.
We propose to calculate the tangent space at I. We replace A by I +X. Then
we have
t
(I + X)(I + X) = I, i.e.,
t
X + X +
t
XX = 0. And then the linear
terms
t
X +X = 0 dene the tangent space at I.
Another way to look at derivations:
Lemma.
(i) The ring homomorphisms : A
A[]
(
2
)
with (a) = a +
1
(a) are in
bijective correspondence with the set of derivations on A.
(ii) Let X be a variety and A = k[X]. A ring homomorphism : A
k[]
(
2
)
when composed with the natural map from
k[]
(
2
)
to k =
k[]
()
denes a
18
point on the variety X. The set of ring homomorphisms from A to
k[]
(
2
)
written as (a) =
0
(a) +
1
(a) is in bijective correspondence with the
tangent space associated to the point x X dened by
0
.
Proof. (i) If d : A A is a derivation on A, then one can check that the
map : A
A[]

2
given by, a a + d(a) denes a ring homomorphism.
Conversely, if (a) = a +
1
(a) is one such ring homomorphism, then
1
is a
derivation on A.
(ii) Here we note that the map k[X]

k[]
(
2
)

k[]
()
= k is precisely the
evaluation at a point x X. This is precisely the way a point on a variety is
dened. Then rest follows as above.
Proposition. Suppose there exists a morphism : G X X of ane
algebraic varieties. Then there exists a map from T
e
(G) to the set of vector
elds on X. If A is an automorphism of X which commutes with the action
of G, i.e., g(Ax) = A(gx), then (v) is an invariant vector eld on X.
Moreover there exists a map from T
e
(G) T
x
(X) to T
x
(X).
Proof. We have : G X X. This gives rise to a map on the level of
coordinate rings,

: k[X] k[G] k[X]. Let v T

e
(G). By second part in
the previous lemma, we get a ring homomorphism
v
: k[G]
k[]
(
2
)
. The ring
homomorphism

when composed with

v
, gives rise to a vector eld on X as
shown in the commutative diagram below. We dene it to be (v). Thus we
get a map from T
e
(G) to the set of vector elds on X.
k[X]

//
(v)
''
P
P
P
P
P
P
P
P
P
P
P
P
P
k[G] k[X]

k[]
(
2
)
k[X] =
k[X][]
(
2
)
.
Now, if A be an automorphism of X which commutes with the G-action,
then we have following commutative diagram:
GX

X
IdA

_A
GX

X
19
This gives us another commutative diagram:
k[X]

k[G] k[X]

v

k[X][]
(
2
)
A

_
IdA

_B

k[X]

k[G] k[X]

v

k[X][]
(
2
)
,
where B

is the natural extension of A

to
k[X][]
(
2
)
. Thus we get a vector eld,
which commutes with every G-invariant automorphism of X. Now, consider
X = G and consider the left action of G on itself, then to every tangent vector
at e to G, we get a vector eld on G which is right invariant. Similarly T
e
(G)
could be identied to left invariant vector elds. The vector eld X
v
associated
to a vector v T
e
(G) has X
v
(e) = v.
Now, we want to give a map from T
e
(G) T
x
(X) T
x
(X). For a
v T
e
(G), we have a vector eld (v) : k[X]
k[X][
1
]
(
2
1
)
. Consider the map
given by
k[G] k[X]
k[
1
]
(
2
1
)

k[
2
]
(
2
2
)

k[]
(
2
)
where the ring homomorphisms
,
are parametrised by the maps
1

and
2
. So, we have
k[X]
//
))
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
k[G] k[X]
//
k[
1
]
(
2
1
)

k[
2
]
(
2
2
)

k[]

Corollary. The space T

e
(G) is isomorphic to the space of left G-invariant
vector elds on G.
Thus T
e
(G) acquires a Lie algebra structure.
Theorem. (Lie algebra of a Lie group). If G is a Lie group, then the set
of left invariant vector elds forms a nite dimensional vector space g which
is closed under Lie brackets; and is called the Lie algebra associated to the
Lie group G. Moreover, dimg = dimG.
20
Proof. Follows from previous proposition and its corollary.
6. More about Algebraic Groups.
In the next two sections, we shall sketch an outline of the theory of ane
algebraic groups. We shall avoid proving theorems. Our emphasis will be on
exposition. We mainly follow Springers book for the exposition. We also list
some open questions in this area.
We have already noted that an ane algebraic group is a closed subgroup of
GL
n
for some n. We also know that every matrix g GL
n
admits a decomposi-
tion, called as Jordan decomposition, as g = g
s
g
u
, where g
s
is semi-simple (i.e.,
g
s
is diagonalisable over k) and g
u
is unipotent (i.e., every eigenvalue of g
u
is 1).
Moreover g
s
g
u
= g
u
g
s
. If g G GL
n
, then g
s
, g
u
G. This decomposition
is called as the Jordan decomposition in the algebraic group G. More generally,
if : G G

is a homomorphism of algebraic groups then (g)

s
= (g
s
) and
(g)
u
= (g
u
). An element g G is said to be unipotent if g = g
u
.
A unipotent algebraic group is the one in which every element is unipotent.
Example: The simplest unipotent group is G
a
, given by
G
a
=
__
1 x
0 1
_
: x k
_
.
Theorem. Unipotent algebraic groups are precisely the closed subgroups of
upper unitriangular group U
n
, (upto conjugacy), for some n.
A consequence of the above theorem is that for a representation of a unipotent
algebraic group G in GL
n
, there is a non-zero vector in k
n
xed by all of G. And
using this fact, we get the following
Proposition. Let X be an ane variety admitting an action of a unipotent
algebraic group G, then all orbits of G in X are closed.
Classication of unipotent algebraic groups is not well understood. The
unipotent algebraic groups can vary in continuous families, i.e., there exists U(s)
unipotent algebraic group parametrised by a complex number s such that for
s
1
= s
2
, U(s
1
)

= U(s
2
).
A solvable (resp. nilpotent) algebraic group is an algebraic group which is
solvable (resp. nilpotent) as an abstract group. If H and K are subgroups of an
algebraic group G, [H, K] denotes the closed subgroup of G generated by the
set {xyx
1
y
1
: x H, y K}.
21
A unipotent group is a closed subgroup of U
n
, and hence is solvable (in fact,
it is nilpotent). More precisely, for a unipotent group U, we dene U
1
= U and
U
i
= [U, U
i1
]. Then U
i
/U
i+1
is an Abelian group, isomorphic to G
d(i)
a
for some
integer d(i).
Question. Is the set of unipotent groups, with given integers d(i), a con-
nected family? Is it an algebraic variety?
Examples in Riemann surfaces suggest that this family is not an algebraic
variety, but a quotient of a variety. More concretely, one could ask following
Question. Classify integers d(i) such that this family is positive dimensional.
The theory of unipotent algebraic groups is also closely connected with clas-
sication of p-groups which is yet an unsolved (unsolvable?) problem.
The subgroup U
n
(F
p
) of GL
n
(F
p
) consisting of all upper triangular unipotent
matrices is a Sylow-p subgroup of GL
n
(F
p
). Therefore, just like unipotent alge-
braic groups, every p-subgroup of GL
n
(F
p
) is also a subgroup of upper triangular
unipotent matrices, upto conjugacy.
Exercise 8. Prove that for n < p, there exists a bijective correspondence
between subgroups of U
n
(F
p
) and connected algebraic subgroups of U
n
dened
over F
p
.
(Hint: Associating a Lie algebra to an abstract unipotent group.)
Question. H

(U
n
(F
p
), F
p
) is an algebra over F
p
. Construct this algebra out
of information coming from H

(u
n
, C), where u
n
is the Lie algebra of upper
triangular unipotent matrices.
Theorem. (Lie-Kolchin). A connected solvable algebraic group is a sub-
group of upper triangular matrices (upto conjugacy).
Corollary. If G is a connected solvable algebraic group, then [G, G] is nilpo-
tent.
Proposition. Let G be a connected nilpotent algebraic group.
(i) The sets G
s
, G
u
of semi-simple and unipotent elements (resp.) are
closed, connected subgroups of G and G
s
is a central torus of G.
(ii) The product map G
s
G
u
G is an isomorphism of algebraic groups.
22
Proposition. For a connected solvable group G, the commutator subgroup
[G, G] is a closed, connected, unipotent, normal subgroup. The set G
u
of
unipotent elements is a closed, connected, nilpotent, normal subgroup of G.
The quotient group G/G
u
is a torus.
In other words, for a solvable group G, there exists an exact sequence
0 G
u
G G
r
m
1.
In fact, this sequence is split, and the reason is the Jordan decomposition!
Another important class of algebraic groups is that of commutative algebraic
groups. A commutative algebraic group is solvable, so all above results of solvable
groups hold for connected commutative groups as well.
Lemma. A connected linear algebraic group G of dimension one is commu-
tative and it is isomorphic to G
a
or G
m
.
A linear algebraic group Gis said to be diagonalisable if it is a closed subgroup
of D
n
, the group of diagonal matrices in GL
n
, for some n.
Lemma. An algebraic group G is diagonalisable if and only if any represen-
tation of G is a direct sum of one dimensional representations.
Theorem. Let G be a diagonalisable group. Then
(i) G is a direct product of a torus and a nite Abelian group of order
prime to p, where p is the characteristic of k;
(ii) G is a torus if and only if it is connected.
Rigidity of diagonalisable groups. Let G and H be diagonalisable groups
and let V be a connected ane variety. Let : V G H be a morphism
such that for any v V , the map x (v, x) denes a homomorphism of
algebraic groups G H. Then (v, x) is independent of v.
For a subgroup H of an algebraic group G, we dene the centraliser and
normaliser of H in G as,
Z
G
(H) := {g G : ghg
1
= h h H};
N
G
(H) := {g G : ghg
1
H h H}.
23
Corollary. If H is a diagonalisable subgroup of G, then N
G
(H)

= Z
G
(H)

and N
G
(H)/Z
G
(H) is nite.
A subgroup P G is called parabolic subgroup if G/P is a projective variety.
Proposition.
(i) If H is a parabolic subgroup of G and K is a closed subgroup, then
H K is a parabolic subgroup of K.
(ii) If H is a parabolic subgroup of K and K is a parabolic subgroup of G,
then H is a parabolic subgroup of G.
(iii) Any closed subgroup of G containing a parabolic subgroup is itself parabolic.
(iv) H is a parabolic subgroup of G if and only if H

is parabolic in G

.
Borels xed point theorem. If G is a connected solvable group operating
on a projective variety X, then G has a xed point.
Proof. We prove this result by using induction on dimG. If dimG = 0,
G = (e). Now let dimG > 0, then H = [G, G] is a closed, connected subgroup
of smaller dimension. Hence by induction hypothesis, the set of xed points of H
in X, say Y , is non-empty and closed in X, thus Y itself is a projective variety.
Since H is normal in G, Y is stable under the G-action. We know that for a
group action, closed orbits always exist. Let y Y such that the orbit Gy is
closed in Y . Look at the isotropy subgroup G
y
of y in G. Then G
y
is closed
subgroup of G and as H G
y
, G
y
is also normal in G. Hence G/G
y
is an ane
variety. But then G/G
y
= Gy, which is projective. Therefore, Gy must be a
point, a xed point of the G-action.
This completes the proof.
A Borel subgroup of G is a maximal connected solvable subgroup of G.
Example: If G = GL
n
, then the subgroup of upper triangular matrices is a Borel
subgroup.
A maximal ag in k
n
is a strictly increasing sequence of subspaces of k
n
,
{0} V
1
V
n
= k
n
. Borel subgroups in GL
n
correspond bijectively
to the set of maximal ags in k
n
. The group GL
n
operates transitively on the
set of maximal ags and isotropy subgroup of the standard ag, {0} {e
1
}
{e
1
, e
2
} {e
1
, . . . , e
n
} is the Borel subgroup described above, i.e., the
subgroup of GL
n
consisting of all upper triangular matrices.
24
Theorem.
(i) All Borel subgroups are conjugates.
(ii) Every Borel subgroup of G is a parabolic subgroup.
(iii) A subgroup of G is parabolic if and only if it contains a Borel subgroup.
(iv) Every element of G lies in a Borel subgroup. In other words, union of
Borel subgroups cover whole of G.
Proposition. Let B be a Borel subgroup of an algebraic group G. Then,
(i) N(B) = B.
(ii) Z(B) = Z(G).
(iii) Any subgroup of G which contains a Borel is always connected.
7. Structure Theory of Reductive Groups.
Here k is a eld, not necessarily algebraically closed.
The unipotent radical of an algebraic group G is the maximal normal unipo-
tent subgroup of G.
Such a subgroup always exists. If U
1
, U
2
are two normal unipotent subgroups
of G, then it can be easily checked that U
1
U
2
is again a normal unipotent
subgroup. Then by dimension argument, one must get a maximal subgroup
of G with these properties. The unipotent radical of G is denoted by R
u
(G).
Example: If G is the subgroup of GL
2
consisting of all upper triangular matrices,
then R
u
(G) is precisely the subgroup of matrices with diagonal entries both equal
to 1.
The radical of a group G is the maximal connected normal solvable subgroup
of G and it is denoted by R(G). One can prove existence of such a subgroup by
similar reasoning as above. Also, R
u
(G) = R(G)
u
. Example: If G = GL
n
, then
R(G) is the subgroup of scalar matrices.
An algebraic group G GL
n
is called reductive if R
u
(G) is trivial. Example:
GL
n
.
An algebraic group Gis called simple if it is non-Abelian and has no connected
normal subgroup other than G and (e). Example: SL
n
.
25
An algebraic group G is called semi-simple if it is an almost product of simple
groups, i.e., G = G
1
G
n
, G
i
G
j
Z(G) for i = j, and all G
i
commute with
each other. Equivalently, an algebraic group G is called semi-simple if R(G) is
trivial. Clearly, every simple algebraic group is semi-simple and every semi-simple
algebraic group is reductive.
Theorem. Any reductive group is upto a center, an almost product of simple
groups.
Theorem. If k = k, then there exists bijective correspondence between simple
groups over k and compact simple groups over R.
This bijection is achieved via the root systems.
We recall that a torus is a diagonalisable commutative group. A torus T G
is called a maximal torus if it is of maximum possible dimension. Example: The
subgroup of diagonal matrices in GL
n
is a maximal torus in GL
n
.
Theorem. Any two maximal tori of an algebraic group are conjugates over
k.
In particular, dimension of a maximal torus in G is a xed number. We dene
it to be the rank of the group G.
For a maximal torus T G, we dene Weyl group of G with respect to T
as N(T)/T. It is independent of the maximal torus if k = k, in that case we
simply denote it by W. Example: The rank of GL
n
is n.
Bruhat Decomposition. Let G be a reductive group, then for a xed Borel
subgroup B,
G =

wW
BwB.
There is the notion of a BN-pair, where N = N(T), for T B, a maximal
torus. These notions were developed by Tits, based on Chevalleys work, to prove
that G(k) is abstractly simple group when G is simple.
Theorem. Let G be a simple algebraic group dened over k. Suppose that
G is split, i.e., G contains a maximal torus which is isomorphic to G
d
m
over
k, where d is the rank of G. Then except for a few exceptions G(k)/Z is a
simple abstract group.
The exceptions are G = SL
2
(F
2
) and SL
2
(F
3
).
26
There is a notion of quasi-split groups. These are the groups that contain a
Borel subgroup dened over k.
Theorem. (Lang). Any reductive algebraic group over a nite eld is
quasi-split.
Steinberg proved analogue of Chevalleys theorem for quasi-split groups.
The group SO(p, q) is quasi-split if and only if |p q| 2.
The group U(p, q) is quasi-split if and only if |p q| 1.
Kneser-Tits Conjecture. Let G be a simply connected algebraic group.
If there exists a map from G
m
to G (i.e., G is isotropic), then G(k)/Z is
simple.
Platonov proved that this conjecture is false in general, but true over global
and local elds.
An algebraic group is called as anisotropic if it does not contain any subgroup
isomorphic to G
a
or G
m
. Example: SO(2n, R), another example is SL
1
(D), the
group of norm 1 elements in a division algebra D.
Question. Which SL
1
(D) are simple?
Platonov-Margulis Conjecture. Over a number eld k, SL
1
(D) is simple
if and only if D k
v
is never a division algebra for a non-Archimedean
valuation v of k.
The question of classifying algebraic groups over general elds forms a part
of Galois cohomology. This is known only for number elds or their completions.
Theorem. Let k be either a number eld or its completion. The tori over k
of dimension n are in bijective correspondence with isomorphism classes of
Gal(k/k)-modules, free over Z of rank n.
Example: The tori over R are the following
T

= (S
1
)
r
1
(C

)
r
2
(R

)
r
3
.
This amounts to classifying matrices of order 2 in GL
n
(Z).
{e e} R

{e e} S
1
27
_
e
1
e
2
e
2
e
1
_
C

Any invariant in GL
n
(Z) is a direct sum of these.
Rationality question. The question is whether k[G] is unirational/rational
for an algebraic group G, i.e., whether k[G] is a purely transcendental exten-
sion of k or contained inside one such.
The answer for above question is yes for groups of type B
n
, C
n
, D
n
.
Cayley Transform: Let G = SO(n) = {
t
AA = I}. Let X be a skew symmetric
matrix, i.e.,
t
A + A = 0. If no eigenvalue of such an X is 1, then I X is
invertible, and it can be checked that (I +X)(I X)
1
belongs to SO(n). This
gives a birational map from the Lie algebra of SO(n) to SO(n), proving the
rationality of SO(n).
8. Galois Cohomology of Classical Groups
Let G be a group, and A another group on which G acts via group automor-
phisms: g(ab) = g(a)g(b). Dene,
A
G
= {a A|g a = a g G} = H
0
(G, A).
If A is commutative, then H
i
(G, A) are dened for all i 0, and these are
abelian groups.
If A is non-commutative, then usually only H
1
(G, A) is dened, and it is a
pointed set:
H
1
(G, A) =
{ : G A|(g
1
g
2
) = (g
1
) g
1
(g
2
)}
{
a
where
a
(g) = a
1
(g)g(a)}
.
If
0 A B C 0,
is an exact sequence of groups, then there exists a long exact sequence:
0 H
0
(G, A) H
0
(G, B) H
0
(G, C)
H
1
(G, A) H
1
(G, B) H
1
(G, C)
This is a long exact sequence of pointed sets: inverse image of the base point
= image of the previous map.
28
If
0 A B C 0,
is an exact sequence of abelian groups, then
0 H
0
(G, A) H
0
(G, B) H
0
(G, C)
H
1
(G, A) H
1
(G, B) H
1
(G, C)
is a long exact sequence of abelian groups.
Let E be an algebraic group over a eld k, for instance the groups like
G
m
, G
a
, GL
n
,
SO
n
, Sp
2n
,
n
, Z/n. Then it makes sense to talk of E(K) for K any eld
extension of k, or in fact any algebra containing k.
If Ais a commutative algebraic group over k, one can talk about H
i
(Gal(K/k), A(K))
for all nite Galois extensions K of k, whereas if A is noncommutative, we
can talk only about the set H
1
(Gal(K/k), A(K)).
Dene
H
i
(k, A) = Lim
K
H
i
(Gal(K/k), A(K)),
direct limit taken over all nite Galois extensions K of k.
The group/set H
i
(k, A) is called the i-th Galois cohomology of A over k.
Example :
1. H
i
(k, G
a
) = 0 for all i 1. This is a consequence of the normal basis
theorem.
2. H
1
(k, G
m
) = 0. This is the so-called Hilberts theorem 90.
3. H
1
(k, GL
n
) = 0.
4. H
2
(k, G
m
) is isomorphic to the Brauer group of k dened using central
simple algebras.
5. H
1
(k, O(q)) is in bijective correspondence wth the isomorphism classes
of quadratic spaces over k.
6. H
1
(k, SO(q)) is in bijective correspondence wth the isomorphism classes
of quadratic spaces over k with a given discriminant.
7. H
1
(k, Sp
2n
) = 1.
29
Example (Kummer sequence): Dene
n
= {x

k

|x
n
= 1}. This is a
Galois module which sits in the following exact sequence of Galois modules:
1
n

1.
The associated Galois cohomology sequence:
1
n
(k) k

n
k

H
1
(Gal,
n
) H
1
(Gal,

) = 1.
Thus H
1
(k,
n
)

= k

/k
n
.
Similarly it can be deduced that H
2
(k,
n
) is isomorphic to the n torsion
in the Brauer group of k.
Example (Spin group) To any (non-degenerate) quadratic form q over
k, we have the special orthogonal group SO(q), and also a certain 2-fold
covering of SO(q), called the Spin group associated to the quadratic form q.
We have the following exact sequence of algebraic groups:
1 Z/2 Spin(q) SO(q) 1.
The associated Galois cohomology exact sequence is:
1 Z/2 Spin(q)(k) SO(q)(k) k

/k
2
H
1
(k, Spin(q)) H
1
(k, SO(q)) H
2
(k, Z/2).
The mapping SO(q)(k) k

/k
2
is called the reduced norm mapping.
The mapping H
1
(k, SO(q)) H
2
(k, Z/2) = Br
2
(k) corresponds to send-
ing a quadratic form q
x
to w
2
(q
x
)w
2
(q) where w
2
is the Hasse-Witt invariant
of a quadratic form.
The basic theorem which is the reason for the enormous usefulness of
Galois cohomology is the following.
Theorem 1 1. The set H
1
(Gal(K/k), Aut(G)(K)) is in bijective corre-
spondence with the set of isomorphism classes of forms of G over k,
i.e., sets of isomorphism classes of groups E over k such that G

= E
over K.
30
2. (Weil Descent) More generally, for any algebraic variety X over k, the
set
H
1
(Gal(K/k), (Aut(X))(K)) is in bijective correspondence with the set
of isomorphism classes of forms of X over k, i.e., sets of isomor-
phism classes of varieties Y over k such that X

= Y over K.
3. Let V be a vector space over k. Let
1
,
2
, ,
r
be certain tensors in
V
a
V
b
. Let G be the subgroup of the automorphism group of V
xing the tensors
i
. Then H
1
(Gal(K/k), G(K)) is in bijective corre-
spondence with the set of isomorphism classes of tensors
1
, ,
r
in
V
a
V
b
such that there exists g Aut(V )(K) such that g
i
=
i
.
Examples :
1. The set H
1
(k, O(q)) is in bijective correspondence with the set of quadratic
forms over k.
2. H
1
(k, Sp
n
) = (1).
3. By the Skolem-Noether theorem, the automorphism group of the al-
gebra M
n
(k) is PGL
n
(k). Hence, H
1
(Gal(K/k), PGL
n
(K)) is the set
of isomorphism classes of central simple algebras of dimension n
2
over
k. Dene Br
K
k
= Ker{Br
k
Br
K
} obtained by sending A to A K.
From the exact sequence,
1 G
m
GL
n
PGL
n
1,
it follows that there is an injection of H
1
(Gal(K/k), PGL
n
(K)) into
H
2
(Gal(K/k), K

). The two maps dened here can be combined to

produce a map from Br
K
k
to H
2
(Gal(K/k), K

) which can be proved

to be an isomorphism.
4. The set of conjugacy classes of maximal tori in a reductive group G with
a maximal torus T, normaliser N(T), is H
1
(k, N(T)). This implies that
the tori in a split reductive algebraic group over a nite eld k are in
bijective correspondence with the conjugacy classes in the Weyl group:
H
1
(k, T) = 0 H
1
(k, N(T))
H
1
(k, W) H
2
(k, T) = 0.
31
Theorem 2 (Hasse-Minkowski theorem)
1. A quadratic form over Q represents a zero if and only if it represents
a zero in Q
p
for all p, and also in R.
2. Two quadratic forms over Q are equivalent if and only if they are equiv-
alent at all the places of Q.
We can interpret the Hasse-Minkowski theorem using Galois cohomology
and then the statement naturally generalises for other reductive groups.
Theorem 3 The natural mapping from H
1
(k, O(q)) to

v
H
1
(k
v
, O(q
v
)) is
one-to-one, where the product is taken over all the places of k.
This brings us to the following conjecture, called the Hasse principle,
proved by Kneser, Harder, Chunousov.
Conjecture 1 Let G be a semi-simple simply connected algebraic group over
a number eld k, then the natural mapping from H
1
(k, G) to

v
H
1
(k
v
, G)
is one-to-one, where the product is taken over all the places of k.
It is a consequence of the Hasse principle that if the number eld has
no real places, then for a semi-simple simply connected group, H
1
(k, G) =
1. Serre conjectured that this vanishing statement is true for all elds of
cohomological dimension 2. This was proved by Eva-Bayer and Parimala for
all classical groups.
32